Sony FDR-AX100

[Dave Blackhurst]

Originally Posted by Adriano Moroni

Out of curiosity: in general with a 4k camera if I like to get a good HD file, is it better to rec in XAVC and to convert it into HD files or directly to rec in HD? In other words what are the differences of image quality (video) between a HD converted file and a HD file (with same camera)? Thanks

This will be one of the questions that will be hard to answer before "hands on"... the 1080/60p will also be at a high bitrate, so there should be some noticeable improvements in 1080p footage shot that way, and I would expect in SOME situations involving high motion, the 60p could be a "better" choice, but in others the 4K will be "better"...

I could be completely wrong when the camera becomes available and we can test actual performance, but I think the above makes some sense... and it's one reason I'm a little less concerned about the 30p limitation of the 4K, I expect the 1080 60p will be pretty impressive, and make for a versatile camera overall.

[James Hobert]

I wish this little guy wasn't so...little. It's shaping up to be amazing for the price but I'd feel odd using a little handycam for any gigs, despite the fact that it's almost universally acceptable now to shoot things with little DSLRs.

I just wish this sensor was dropped in the Z100 or some other pro-cam. Always so close...

[Ron Evans]

[QUOTE=James Hobert;1835278

I just wish this sensor was dropped in the Z100 or some other pro-cam. Always so close...[/QUOTE]

Then it would need a bigger lens than is currently on the Z100 or AX1. I would like that too if the features of the FDR-AX1 and PXW-Z100 stayed. Maybe we will see that in the future but then expect the price to be more than these cameras not just a little more than $2000 !!! I still think we will see an EA50 4K version as that would make sense in the lineup for Sony. Sort of a large sensor interchangable lens addition to the FDR-AX1 at much the same price I expect.

Ron Evans

[John McCully]

My RX10 is 50Hz/60Hz (in historical terms NTSC/PAL) switchable. I shoot at 60Hz because when playing 50Hz files on a computer monitor I observe motion artifacts with 50Hz files. This happens with my EX1 also.

I prefer 30p vs 25p also.

I wonder if the FDR-AX100 is also switchable like the RX10 version sold in New Zealand, a PAL country. I note at the Sony UK site the following options: XAVC S 4K :3840×2160/25P, 24P, "XAVC S HD :1920x1080/50P, 25P, 24P", AVCHD:1920x1080/50p(PS),24p(FX,FH),50i(FX,FH), 1440x1080/50i(HQ,LP), MP4: 1280x720 25p

And at the Sony USA site: XAVC-S 4K : 3840×2160/30P, 24P; XAVC-S HD : 1920x1080/60P, 30P, 24P;AVCHD:1920x1080/60p(PS),24p(FX,FH),60i(FX,FH), 1440x1080/60i(HQ,LP); MP4: 1280x720 30p

All of which leads me to believe the FDR-AX100 will not be switchable. That means B&H will get my order and not Sony New Zealand. No mention of the FDR-AX100 at either the New Zealand or the Australian site that I can find.

I wonder why the RX10 is switchable and how and why Sony decide to include switchability, or not?

[Dave Blackhurst]

Well John, you won the coin flip this time, the US market RX10's aren't' switchable... go figure.

[Glen Vandermolen]

Originally Posted by Ron Evans

Then it would need a bigger lens than is currently on the Z100 or AX1. I would like that too if the features of the FDR-AX1 and PXW-Z100 stayed. Maybe we will see that in the future but then expect the price to be more than these cameras not just a little more than $2000 !!! I still think we will see an EA50 4K version as that would make sense in the lineup for Sony. Sort of a large sensor interchangable lens addition to the FDR-AX1 at much the same price I expect.

Ron Evans

Would it need a bigger lens...or have a lens at roughly the current size, but only a 10x zoom. I think many would swap the extra zoom range for a larger sensor.

[Ron Evans]

I was thinking of a 20x zoom, F1.6 etc, which would be bigger for a 1" sensor I think. Of course one could just put the FDR-AX100 guts in the larger body. If that would have the encoders of the larger cameras that would be great.

Ron Evans

[Peter McLennan]

Just finished this entire thread in one read. :) Nice to read some educated discussion rather than the usual fanboi drivel.

I'm in the market for new cameras. My two TRV 900's just don't cut it any more.

This NAB will be my tenth or so. Are you guys plotting a meetup? I'd love to buy beer and listen.

[Peter Siamidis]

Originally Posted by Ron Evans

So the real world performance may only be 1/3 to a stop faster than the AX1.

One question, many others cameras in that price range do line skipping to downscale the image to the desired resolution when used for video. Compared to that the AX100 as I understand it actually reads the entire sensor then downscales after that. Does that mean it would perform better in low light since while other cameras effectively throw away much of the light they have gathered before sending the image to processing, the AX100 uses all gathered light from the sensor and then processes that? I think Sony calls is 'Direct Pixel Readout" or something like that in their jargon, but I believe all it means is that they don't line skip and instead use all data collected by the sensor. I don't think the AX1 has that, so maybe that would make a difference in low light performance when compared to the AX100?

[Anthony Lelli]

old marketing, will they ever change? no 60p? really? but in the stills they were forced to lift all the limitations and give the same for all the models, like all the others do with stills): but with video they are back to the 1980 marketing. Guys it's up to us now: we have to let them know that this old marketing is in the past. or they think that we are stupid or something.

[Ron Evans]

The AX1 has 8M pixels exactly correct for 3840x2160 so has to use them all. Don't be confused by the sensor specs as it is not a 16x9 sensor coming from a still camera. The crop for the lens in the AX1 or PXW-Z100 only addresses a 1/3" crop of the full sensor. AX100 has 14M ( not sure if this is 16x9 either: Just checked and it is the 16x9 )so does have to downscale a little to the 8.3M needed and how this affects the low light performance. I tried to account for this in my very simple calculation. We will have to wait to see what the real world performance is like between the two cameras. My guess is at full wide angle there may not be much difference and the AX100 will have a shallower depth of field.

Ron Evans

[Cliff Totten]

I have been looking at this "cat video" again on a 27 inch iMac (2560 x 1440, not it's not 4K)

This image Jell-O's badly. The more I look at it, the more I cringe at the motion artifacts. Hopefully, this is just a prototype problem.

If it's not a prototype issue and all the final cameras have this...than this explains why it was allowed to have full manual controls for this low of a price. Sony gave this little guy everything and that shocked me and I couldn't figure out why that did that for only $2000.

If rolling shutter "jell-o" is going to be the AX100's big Achilles Heel, than Sony can certainly load up and "de-cripple" all the rest of the functions without any sales cannibalism worries.

If the next Sony 4k camera that is priced above the AX100 has significantly less skew than it wont feel any competition from the AX100. (that alone is enough to protect a higher and more profitable model's sales)

If the AX100 is deliberately built this way than it's a brilliant marketing strategy on Sony's part.

It's very interesting that the RX10 does not at all show that amount of skew.

CT

[Dave Blackhurst]

@ Anthony - the AX100 has higher bitrate 1080 60p than the RX10, so it's the RX10 that is "crippled" in that sense... and IMO it COULD have done 4K... whatever, I'm enjoying the camera and shooting with it quite a lot, as-is. I'll pick up a slightly used AX100 from someone who expects it to bring them a chilled beer and is "mad" because it doesn't, and I'm sure it will be a good camera.

I know you want to argue that somehow the manufacturers are playing some game and ripping "us" off, but EVERY camera has compromises, EVERY technology has limits, and EVERY company, if they hope to keep supplying new and innovative products, has to make money once in a while! Add to that the fast rate of change in ANY electronic product, and yes, "next years product" will most likely be "better" than "last year's product", making it "seem" like they held stuff back or something (Apple has this down to an art...)

Yes, companies make design decisions, things like a "slow" zoom that many complain about as being TOO slow, but it prevents amateurs from crash zooming, something that probably is a good design call in a "consumer" product... perhaps they realize that putting a 4K/60p product on the market without a reasonably priced and readily available MEDIA to record to would cause a "few" problems, so they stuck to a more feasible 30p at a lower bitrate.

If you transported some video guy from "5 years ago", and showed him the new toys, he'd be mad if you sent him back in time... OK, "10 years ago" guy would go "whut, when did the HC1 get 4K?!?". He'd still be blown away by the image quality....




@Cliff - just spitballin', but maybe the RX10 is reading/processing 1/4 of the information for a 1080 "frame" - so it probably is taking relatively longer to do a full read of the sensor to the processor to produce the 4K image?

One of the reasons to get the AX100 over the CX900 is to see where the "issues" may arise with 4K. Inherently there's a lot more data being tossed around that must be dealt with and accounted for somehow. I'm quite certain there will be bumps in the road....

The AX100 is a "first", and we'll have to see where it shines, and undoubtedly will run into places where it falls a bit short. Some of the early "test video" from the RX10 was actually pretty badly skewed and lots of "motion artifacts" due to bad camera operators/"reviewers".

I'm generally skeptical of "first tests" where swinging the camera wildly in a way NO ONE would do in real life, except to "prove a point" (that you CAN swing a camera wildly and get bad results)... is regarded as "testing methodology". If it's done to show that rolling shutter exists, well, slap me with a fish, I did not know that! Oh wait... we did... but of course there's no "standardized" test of how far and how fast the camera is swung wildly, so the "example" proves... you can swing a camera wildly and get bad results...

When the camera becomes more widely available, then we will see what it can do when properly handled...

[Ken Ross]

Color me as unimpressed by the wild swinging and resulting jello. I guess if that test bothers some, the AX100 is not for them. If I saw this jello throughout the video and in shooting scenarios like I'm accustomed to, then I would be concerned.

But I guess I'm the oddball and tend not to swing my camera to & fro, back & forth in an attempt to nauseate my audience. No, I tend to shoot more like the tester did prior to the frenzied "rockin & rollin' where no jello was visible. :)

As for the lack of 60p, yup, I guess I'll be 'stupid or something' and buy the damn thing! ;)

[Troy Lamont]

Originally Posted by Cliff Totten

I have been looking at this "cat video" again on a 27 inch iMac (2560 x 1440, not it's not 4K)

This image Jell-O's badly. The more I look at it, the more I cringe at the motion artifacts. Hopefully, this is just a prototype problem.

Are you looking at the Vimeo, YouTube or download version? There are problems inherent with all three which I'm sure you're well aware of, but keep in mind that if you're looking at the 1080p download from Vimeo that it's being upconverted to WQHD by your graphics card.

I'm not doubting that you're seeing what you're seeing but we're looking at and analyzing converted conversions. ;-)

Raw 4K files will be the tell tale, only 11 days left. He also confirmed that his camcorder is a preproduction model yesterday on his Vimeo page and we don't know if he engaged the OIS etc.